These two decision biases show how people misread chance and commitment. They often feel rational in the moment, yet they systematically push choices away from evidence, probabilities, and long-term goals.

Gambler’s Fallacy

Core idea: misjudging randomness

People often expect random events to “balance out” in the short run, even when each event is independent.

A key psychological driver is pattern perception: humans are motivated to find order, so truly random sequences can feel “wrong” if they contain clusters or streaks. This bias is strongest when outcomes are vivid, frequent, and emotionally charged (for example, wagering contexts), and it can promote riskier bets and persistence.

Independence and “due” thinking

The fallacy commonly appears when:

• A person treats independent events (like coin flips) as if they influence each other.

• A person confuses the law of large numbers with short-run expectations (long-run averages are misapplied to small samples).

• A person uses intuitive “fairness” beliefs (randomness is expected to look evenly mixed).

Practical consequences for decision-making include:

• Escalating commitment after losses because a win feels imminent.

• Overconfidence in predictions based on recent outcomes rather than base rates.

• Poor probability judgments in games of chance and in everyday forecasting (where people incorrectly assume trends must reverse).

Distinguishing from related errors (without relying on them)

Not all streak-based thinking is gambler’s fallacy. If a process is not independent (for example, fatigue affecting performance), expecting change can be reasonable. Gambler’s fallacy specifically involves unchanged underlying probabilities.

Sunk-Cost Fallacy

Core idea: past costs distort future choices

When people have already invested resources, they may continue investing to “justify” what they have spent, even when stopping would be wiser.

A flow chart that contrasts a normatively rational decision rule (focus on future costs/benefits) with the sunk-cost fallacy (continuing because of prior investment). The diagram makes explicit how irrecoverable past costs can hijack the decision criterion, producing escalation of commitment. Used as a study aid, it reinforces that sunk costs change feelings, not probabilities or payoffs. Source

This bias persists because prior investments feel psychologically meaningful, producing loss aversion and self-evaluative concerns (for example, not wanting to feel wasteful). The result is a focus on recovering past costs rather than maximising future outcomes.

Sunk costs can include:

• Money already spent

• Time already invested

• Effort already exerted

• Reputation or public commitment already made

Why sunk costs feel compelling

Several mechanisms maintain the bias:

• Cognitive dissonance: stopping can imply the original choice was mistaken, creating discomfort.

• Self-justification: persisting protects self-image (“I’m not the kind of person who quits”).

• Escalation of commitment: each additional investment increases the felt need to continue.

• Emotional accounting: people treat past effort as something that must “pay off,” even when it cannot affect future payoffs.

Decision-making consequences include:

• Staying in failing plans, projects, or obligations longer than evidence supports.

• Taking extra risks to “get even,” which often increases losses.

• Ignoring opportunity costs (what could be gained by switching to a better alternative).

Sound decision rule (future-focused)

A psychologically informed approach is to separate irrecoverable past costs from future expected outcomes:

• Ask what you would choose if you were deciding fresh today.

• Prioritise forward-looking information (likely benefits, likely harms, and feasible alternatives).

• Treat stopping as an information-based update, not a moral failure.

How these fallacies interfere with sound decision-making

Both biases pull attention away from relevant data:

• Gambler’s fallacy: overweights recent sequences and underweights true probabilities.

• Sunk-cost fallacy: overweights past investments and underweights future consequences.

In both cases, the mind substitutes a compelling narrative (streaks must reverse; investments must be redeemed) for probabilistic and goal-directed reasoning.